Using different versions of CellProfiler

Hi, I have a problem using two different versions of Cell Profiler: first I utilized Cell Profiler 1.0 on some images, then when I utilized Cell Profiler 2.0 on the same images with the same pipeline I obtained very different results. Moreover I obtain different results also on some examples I downloaded from the website. How can I resolve this problem? do you think that one version is more reliable than the other one?
Thank you!

Hi,

Could you upload a pipeline plus a sample of the images which gave you differing results?
-Mark

And in which Example pipelines do you find differences?

We are very keen to keep CP2 measurements consistent with CP1, of course. We find that CP2 is at least as reliable as CP1, as we are doing more rigorous unit testing now, and have fixed many bugs in the code rewrite, though some discrepancy may have slipped through.

Thanks,
David

Hy, the differences regard some measurements, like the integrated intensity of the Nuclei. I think that you could start trying with the example pipeline ExampleHumanImages, the count of nuclei is the same but the measurements are different. Eventually, we will attach you our pipeline and images later. Thanks a lot :smiley: and have a nice weekend
Lucia, Imma and AnnaMaria

Hi all,

I took a look at the ExampleHuman data and I think I know what you’re seeing. Between CP 1.0 and CP 2.0, the row ordering of the objects is not the same, so you may have been looking at different objects. If you order the nuclei by their location, then you should see that the measurements are similar between the two versions for each object (they should match to several decimal places).

Regards,
-Mark